Skip to content
Navigation Menu
Sign in
Products
Platform
GitHub
Repository hosting, code review, and collaboration
Copilot
AI coding agent for tasks, projects, and automated work
Build & Ship
Actions
CI/CD and automation
Code Quality
Quality analysis, fixes, and enforcement
Codespaces
Cloud development environments
Secure
Code Security
Vulnerability detection and remediation
Secret Protection
Secret scanning and leak prevention
All Features
Developers
Learn
Documentation
Training
Certifications
Build
Build with Copilot
Customize Copilot
Integrations
Explore
Trending repos
GitHub Sponsors
Developer articles
Changelog
Changelog
Your source for the latest features, improvements, and fixes across GitHub products, all in one place.
Businesses
By Stage
Enterprises
Team
Startups
Nonprofits
By Industry
Healthcare
Financial services
Manufacturing
Government
See all industries
Evaluate GitHub
Why GitHub
Customer Stories
Business insights
Trust Center
Add-ons
GitHub Advanced Security
Enterprise-grade code security
Copilot for business
Enterprise-grade AI features
Premium Support
Enterprise-grade 24/7 support
Resources
Learn & Connect
Documentation
Resource center
Blog
Events & webinars
Community Forum
Customer Stories
Integrations
Support
Customer support
Training
GitHub Status
Company
About
Partners
Changelog
GitHub Universe
GitHub Universe
The event for developers, by developers — talks, workshops, and what’s next for GitHub.
Pricing
Search
/
Sign in
Sign up
Appearance settings
You signed in with another tab or window.
Reload
to refresh your session.
You signed out in another tab or window.
Reload
to refresh your session.
You switched accounts on another tab or window.
Reload
to refresh your session.
Dismiss alert
{{ message }}
SemiAnalysisAI
InferenceX
Repository navigation
Code
Issues
8
(8)
Pull requests
153
(153)
Discussions
Actions
Projects
Security and quality
Insights
More
items
All pull requests
Labels
64
(64)
Milestones
5
(5)
New pull request
Search pull requests
is
:
pr
state
:
open
is:pr state:open
Clear filter
Search
Pull requests
Open
153
(153)
Closed
4,536
(4,536)
Author
Label
Projects
Milestones
Reviews
Assignee
Sort by
Newest
descending
Sort…
Sort by Newest descending
More items
Comfortable display density
Compact display density
musespark mi355x: append TP4/EP4 concurrency 58 / musespark mi355x:追加 TP4/EP4 并发 58
#5155
·
dorado269
opened
Nov 20, 2026
·
4 of 11 tasks
·
·
1
[NV] MiniMax-M3.1-Flash on GB300: FP4 Dynamo vLLM — DSpark agg and disagg AgentX points on cu135 image / [NV] MiniMax-M3.1-Flash 在 GB300 上:FP4 Dynamo vLLM — 基于 cu135 镜像的 DSpark 聚合与分离式 AgentX 测试点
#5153
·
RohitNagraj
opened
Nov 20, 2026
Collaborator
·
3 of 11 tasks
·
·
1
[AMD][MI325X] DSv4.1 Flash vLLM: dense decode CUDA graphs at TP4 c16 on the f5d93296 nightly / [AMD][MI325X] DSv4.1 Flash vLLM:在 f5d93296 nightly 上为 TP4 c16 启用密集 decode CUDA graph
full-sweep-fail-fast
#5152
·
ZhengGong-amd
opened
Nov 20, 2026
Collaborator
·
5 of 11 tasks
·
·
4
[AMD][AgentX] DeepSeek-V4.1-Flash MI355X vLLM: rerun the whole curve on the 11-19 nightly / [AMD][AgentX] DeepSeek-V4.1-Flash MI355X vLLM:在 11-19 nightly 上重跑整条曲线
full-sweep-fail-fast
#5151
·
chunfangamd
opened
Nov 20, 2026
Collaborator
·
6 of 11 tasks
·
·
4
Run GLM-5.3-Flash MI300X/MI325X block-FP8 linears on aiter's CK GEMM / GLM-5.3-Flash MI300X/MI325X 的 block-FP8 线性层改用 aiter CK GEMM
full-sweep-fail-fast
#5150
·
fengshaoyi-amd
opened
Nov 20, 2026
Collaborator
·
6 of 11 tasks
·
·
1
[K3.1-Pro] Add GB300 FP4 disaggregated llmd-vllm AgentX configs / [K3.1-Pro] 新增 GB300 FP4 分离式 llmd-vllm AgentX 配置
non-canary-full-sweep-enabled
#5149
·
cquil11
opened
Nov 19, 2026
Collaborator
·
·
6
[NV] GLM-5.3 on GB300: NVFP4 Dynamo TensorRT-LLM — Add disaggregated AgentX / [NV] GLM-5.3 在 GB300 上:NVFP4 Dynamo TensorRT-LLM — 新增 AgentX 分离式配置
non-canary-full-sweep-enabled
#5148
·
RohitNagraj
opened
Nov 19, 2026
Collaborator
·
5 of 11 tasks
·
·
2
perf(qwen3.5): GB300 TRT AgentX on TRT-LLM 1.4.0rc4 with remote KV cache / perf(qwen3.5):GB300 TRT AgentX 改用 TRT-LLM 1.4.0rc4 与远程 KV 缓存
non-canary-full-sweep-enabled
#5147
·
richardhuo-nv
opened
Nov 19, 2026
Collaborator
·
·
5
chore: bump srt-slurm to v2.73.0 / chore:将 srt-slurm 升级到 v2.73.0
#5145
·
seungrokj
opened
Nov 19, 2026
Collaborator
·
·
2
feat(config): add TokenSpeed Kimi K3.1 Pro MXFP4 MI355X AgentX with DSpark / feat(config): 新增 TokenSpeed Kimi K3.1 Pro MXFP4 MI355X AgentX(DSpark)配置
non-canary-full-sweep-enabled
#5144
·
Yu-Zhewen
opened
Nov 19, 2026
Collaborator
·
9 of 11 tasks
·
·
6
[NVIDIA][AgentX] GB300 Muse Spark Open Dynamo+SGLang: retune on the 20261119 nightly / [NVIDIA][AgentX] GB300 Muse Spark Open Dynamo+SGLang:基于 20261119 nightly 重新调优
full-sweep-fail-fast
#5143
·
shyeh25
opened
Nov 19, 2026
Collaborator
·
4 of 11 tasks
·
·
4
[Klaud Cold] Update DeepSeek-V4.1-Flash FP4 GB200 vLLM AgentX DSpark image to cu134-nightly-f5d93296 / [Klaud Cold] 将 DeepSeek-V4.1-Flash FP4 GB200 vLLM AgentX DSpark 镜像更新至 cu134-nightly-f5d93296
full-sweep-fail-fast
klaud-finished
#5141
·
Klaud-Cold
opened
Nov 19, 2026
Collaborator
·
·
11
[AMD][MI355X] Add GLM-5.2 MXFP4 AgentX P/D behind the Infera router (infera-sglang) / [AMD][MI355X] 新增 GLM-5.2 MXFP4 AgentX PD 分离部署(Infera router,infera-sglang)
full-sweep-fail-fast
#5138
·
llying-001
opened
Nov 19, 2026
Collaborator
·
7 of 11 tasks
·
·
4
[AgentX] MiniMax-M3.1-Flash GB300 Dynamo + SGLang: aggregated points on the 1113 nightly, TP2 HiCache to c50 / [AgentX] MiniMax-M3.1-Flash GB300 Dynamo + SGLang:聚合部署点迁移到 1113 nightly,TP2 HiCache 扩展到 c50
full-sweep-fail-fast
#5137
·
nvpohanh
opened
Nov 19, 2026
Collaborator
·
6 of 11 tasks
·
·
10
[Klaud Cold] Update Muse Spark Open FP4 B300 SGLang AgentX DFlash image to nightly-dev-cu13-20261118-ecd3d75a / [Klaud Cold] 将 Muse Spark Open FP4 B300 SGLang AgentX DFlash 镜像更新到 nightly-dev-cu13-20261118-ecd3d75a
full-sweep-fail-fast
klaud-finished
#5134
·
Klaud-Cold
opened
Nov 19, 2026
Collaborator
·
·
16
[AMD][MI35X] Qwen3.8-Flash-Next FP4 SGLang AgentX MTP: opt-in QSA, MXFP4 MoE, HC prefill and GDN verify kernels on the update-monokernel recipe / [AMD][MI35X] Qwen3.8-Flash-Next FP4 SGLang AgentX MTP:在 update-monokernel 配方上启用 QSA、MXFP4 MoE、HC 预填充与 GDN verify 可选内核
all-evals
full-sweep-fail-fast
#5133
·
maeehart
opened
Nov 18, 2026
Collaborator
·
4 of 11 tasks
·
·
2
[AMD][AgentX] run the GLM-5.2 MI355X DCP4 arm on HiCache, on the 20261117 daily image / perf: GLM-5.2 MI355X DCP4 分支改走 HiCache,并统一到 20261117 镜像
agentx
AMD
full-sweep-fail-fast
#5128
·
EricKing626
opened
Nov 18, 2026
Collaborator
·
·
5
[AMD][Qwen3.5] Run MI355X FP8 AgentX on the 20261120 image with the monokernel
full-sweep-fail-fast
#5126
·
yichiche
opened
Nov 18, 2026
Collaborator
·
·
3
[AMD][AgentX] DeepSeek-V4.1-Flash MI355X vLLM: TP4 concurrency 1-48 with running-request caps on the 11-17 nightly / [AMD][AgentX] DeepSeek-V4.1-Flash MI355X vLLM:在 11-17 nightly 上把 TP4 扩展到并发 1-48 并限制同时运行的请求数
#5125
·
fanxingran
opened
Nov 18, 2026
·
5 of 11 tasks
·
·
5
feat(changelog): one perf-changelog file per PR, checked against the config diff / feat(changelog):每个 PR 一个 perf-changelog 文件,并按配置 diff 校验
#5123
·
adibarra
opened
Nov 17, 2026
Collaborator
·
·
7
feat(agentx): add single-node TP8 mono points to the GLM-5.2 MI355X ATOM curve / feat(agentx):GLM-5.2 MI355X ATOM 曲线新增单机 TP8 mono 测试点
full-sweep-enabled
#5120
·
Jasen2201
opened
Nov 17, 2026
Collaborator
·
7 of 11 tasks
·
·
4
perf(amd): Kimi-K3.1 MI355X vLLM AgentX: released DSpark draft, prefix-match-unit 1152, nst 4 at c32/c48 / perf(amd):Kimi-K3.1 MI355X vLLM AgentX:正式版 DSpark 草稿、prefix-match-unit 1152、c32/c48 nst 4
non-canary-full-sweep-enabled
#5119
·
hyukjlee
opened
Nov 17, 2026
Collaborator
·
·
4
[AMD][AgentX] MiniMax-M3.1-Flash MI355X vLLM: append c24, c48, c56, c68 and c72 / [AMD][AgentX] MiniMax-M3.1-Flash MI355X vLLM:新增 c24、c48、c56、c68、c72
#5118
·
benenzhu
opened
Nov 17, 2026
Collaborator
·
·
5
perf(nvidia): GLM-5.3 NVFP4 B300 SGLang AgentX: FlashMLA sparse prefill + TRT-LLM sparse decode / perf(nvidia):GLM-5.3 NVFP4 B300 SGLang AgentX:FlashMLA 稀疏预填充 + TRT-LLM 稀疏解码
non-canary-full-sweep-enabled
#5115
·
Oseltamivir
opened
Nov 17, 2026
Collaborator
·
·
5
[Klaud Cold] Update GLM-5.3-Flash FP8 MI300X SGLang AgentX MTP image to v0.5.23 / [Klaud Cold] 将 GLM-5.3-Flash FP8 MI300X SGLang AgentX MTP 镜像更新至 v0.5.23
full-sweep-fail-fast
klaud-finished
#5107
·
Klaud-Cold
opened
Nov 17, 2026
Collaborator
·
·
7
Previous
1
2
3
4
5
6
7
Next
You can’t perform that action at this time.